IBM renews commitment to enterprise-focused AI by launching $500M fund

IBM recently announced that it would launch the $500-million IBM Enterprise AI Venture Fund as its sole investor. The fund will be aimed at investing in startups focusing on generative AI acceleration and research for the enterprise, regardless of whether they are early-stage or hyper-growth. The fund has a differentiated strategy and model featuring IBM domain experts and a dedicated portfolio development team. The combined model will help IBM achieve its primary goal of allowing startups to develop meaningful partnerships with IBM that will enable them to gain engineering and product knowledge paired with effective marketing strategies, thus creating lasting value for leaders and clients alike.

IBM is one of many companies to ride the latest generative AI-driven wave with watsonx, its proprietary generative AI and data platform for the enterprise. With the launch of the Enterprise AI Venture Fund, IBM looks to expand its partnership AI ecosystem by working together with startups leveraging and building on the watsonx platform. According to Rob Thomas, Senior Vice President, Software and Chief Commercial Officer, generative AI is set to unlock nearly $16 trillion in productivity by 2030, and IBM is planning to capitalize on this prediction by harnessing the potential of generative AI into tangible, positive outcomes for itself and the startups it will invest in. At the same time, the fund's launch also embodies IBM's commitment to responsible AI innovation that results in solutions that enterprises can benefit from.

The Entreprise AI Venture Fund is hardly the first time IBM has invested in AI-focused companies. In August, it announced its participation in Hugging Face's $235M Series D funding round. IBM has also contributed hundreds of open models and datasets on the platform, including the Geospatial Foundation Model developed in partnership with NASA. The Geospatial Foundation Model is the largest geospatial model hosted in Hugging Face and the first-ever open-source foundation model built with NASA. IBM also recently participated in HiddenLayer's Series A round devoted to increasing HiddenLayer's talent headcount, intensifying its marketing efforts, and investing in the company's award-winning Machine Learning Security Platform. HiddenLayer's Series A was the largest one raised by a cybersecurity company focused on AI protection this year.

The news of the Enterprise AI Venture Fund builds on the company's previous experience building partnerships with startups focused on accelerating AI solutions for the enterprise. It also advances IBM's strategy of leveraging both its own and third-party models to foster AI innovation. IBM's own watsonx Granite model series is now generally available to anyone looking to integrate generative AI within business applications and workflows. Moreover, the company also plans to host Meta's Llama-2-70b-chat on the watsonx platform. Given IBM's primary goal of helping ignite innovation as well as furthering partnerships with startups leveraging or building upon watsonx, the availability of a diversity of proprietary and third-party models goes quite well with the launch of the venture fund.